{"id":24855,"date":"2019-08-15T17:14:44","date_gmt":"2019-08-15T17:14:44","guid":{"rendered":"https:\/\/devblogs.microsoft.com\/cppblog\/?p=24855"},"modified":"2019-08-15T17:14:44","modified_gmt":"2019-08-15T17:14:44","slug":"vcpkg-2019-07-update","status":"publish","type":"post","link":"https:\/\/devblogs.microsoft.com\/cppblog\/vcpkg-2019-07-update\/","title":{"rendered":"Vcpkg: 2019.07 Update"},"content":{"rendered":"

The 2019.07 update of vcpkg, a tool that helps you manage C and C++ libraries on Windows, Linux, and macOS, is now available. This update is a summary of the new functionality and improvements made to vcpkg over the past month. Last month was the first time we created a vcpkg release (Vcpkg: 2019.06 Update<\/a>).<\/p>\n

In this post, we will cover caching in Azure Pipelines with vcpkg in addition to many new port and triplet updates, improvements for port contributors, and new documentation. For a full list of this release\u2019s improvements, check out our changelog on GitHub<\/a>.<\/p>\n

Caching in Azure Pipelines with vcpkg<\/h2>\n

The public preview of caching in Azure Pipelines<\/a> is now available, and you can use it with vcpkg!<\/p>\n

You can use pipeline caching to improve build time by allowing previously built and cached vcpkg artifacts (including libraries) to be reused in subsequent runs. This allows you to reduce and avoid the cost to rebuild the same libraries for each build run the same libraries. Caching may be especially useful with vcpkg when you are installing and building the same dependencies (libraries) over and over during your build. That process can often be time-consuming if it involves building large libraries.<\/p>\n

For example, if you have a C++ application that uses SQLite databases, you\u2019ll likely want to use SQLite3 among other libraries. Each time you run a build on your server, you install vcpkg and the sqlite3 library. Without using pipeline caching, this may take some time:<\/p>\n

\"\"<\/p>\n

Now, with Azure Pipelines Caching, we can have a much faster and better experience. One of our community contributors and Microsoft employee, Luca Cappa<\/a>, created a pipeline task to help you use the CacheBeta pipeline task<\/a>. We\u2019ll show you how leveraging his scripts reduced the \u2018run vcpkg\u2019 build step from 2m 26s to just 14s!<\/p>\n

CppBuildTasks Azure DevOps Extension with vcpkg<\/h3>\n

The CppBuildTasks project<\/a> pipeline task does the following:<\/p>\n

\"\"<\/p>\n

So, in the case of SQLite3, vcpkg will be updated and the sqlite3 library will be installed. It is cached in the pipeline such that in subsequent runs you do not need to install and build the sqlite3 library again.<\/p>\n

To get started with CppBuildTasks with Azure Pipelines in your project you can follow the simple CppBuildTasks developer documentation<\/a>.<\/p>\n

Looking at the example, there are a few things to note:<\/p>\n

variables:\r\n    # Exact vcpkg's version to fetch.\u00a0\u00a0 \u00a0\r\n    vcpkgGitRef: 5a3b46e9e2d1aa753917246c2801e50aaabbbccc\r\n\r\n    # Cache the vcpkg's build artifacts.\u00a0 \r\n  - task: CacheBeta@0\u00a0\u00a0 \u00a0\r\n    displayName: Cache vcpkg\u00a0\u00a0 \u00a0\r\n    inputs:\u00a0\u00a0\u00a0\u00a0 \u00a0\r\n      # As 'key' use the content of the response file, vcpkg's commit id and build agent name.\u00a0\u00a0\u00a0\u00a0 \u00a0\r\n      # The key must be one liner, each segment separated by pipe, non-path segments enclosed by\u00a0\u00a0\u00a0\u00a0 \u00a0\r\n      # double quotes.\u00a0\u00a0\u00a0\u00a0 \u00a0\r\n      key: $(Build.SourcesDirectory)\/vcpkg_x64-linux.txt | \"$(vcpkgGitRef)\" | \"$(Agent.Name)\"\u00a0\u00a0\u00a0\u00a0 \u00a0\r\n      path: '$(Build.BinariesDirectory)\/vcpkg'<\/pre>\n